New on LowEndTalk? Please Register and read our Community Rules.
All new Registrations are manually reviewed and approved, so a short delay after registration may occur before your account becomes active.
All new Registrations are manually reviewed and approved, so a short delay after registration may occur before your account becomes active.
Comments
Also how can GreenCloudVPS (mentioned a lot) be reliable for production with these very low prices? https://greencloudvps.com/billing/store/budget-kvm-sale
No because I will lose millions for any downtime.
i still use racknerd for blog kwkwk
LET is great for production if you have a decent RAIV (Redundant Array of Inexpensive VPSes) layer. Not only it's cheaper but it could be more reliable than going with a single "reliable" top vendor like AWS/GCP/Azure.
Why does high price equate high quality?
I've used Hetzner and OVH for production extensively over the years.
Since about maybe 1-2 years MXRoute in production. Before that, RamNode with MailChannels.
It's definitely possible to have stability at low-end pricing.
Now, would I do prod stuff on SolidSeoVPS & Onidel? Probably not, but I did buy their services anyway. There's different levels of production.
Production for me is that more people than me (and friends & family) rely on it to be up and work.
I had a critical incident that lasted 4 days with google cloud.
So yes, some LET providers are quite acceptable for prod.
The most important thing for production is to set up redundancy and a disaster recovery plan.
It does not, but a low price means corners are cut somewhere. High price does not automatically mean high quality, but a low price can only get you so much quality.
Generally speaking, yes.
But this was specifically commented for GreenCloud. I do not see them cutting corners anywhere. Their service has been top notch. Support very responsive and helpful.
The general consensus and I have a few services with them and they have been as reliable as one could imagine. Infact surpassing what one would typically pay for that quality of support and service.
Would you like to test your production service on our infrastructure free of charge for 6 months to 1 year?
It's an extremely kind offer, sadly, all our prod runs in the EU and I don't think I'd want to change that, but again, extremely kind
You're clearly already doing much right, panel is beautiful and service seems stable (my host node says it has an uptime of about 1,5 years)
All you'd have to do to fit into my "production tier" is basically keep what you're doing for longer, I want to see some history of providing a stable services before recommending stuff for prod, but again, what you provide is already clearly above many LET hosts in terms of usability & UI, which is a great sign in determining the longevity of a provider IMO (although not a definite)
Do ping me if/when you go EU
I use several providers here for production services. @1gservers is rock solid (4+ nines), although admittedly not the cheapest here, and I use them to host web sites & email. @OVHcloud is also fine, and I use them for a DNS server and a backup server. I use offerings from other providers here (e.g. Gigahost & Greencloud) for DNS servers, to get geographical diversity without breaking the bank—given the nature of DNS, even 3 nines of uptime/reachability is OK—or for other applications where the performance/price tradeoff is acceptable.
What's nice about LET is that you get a chance to take advantage of some great offers to "kick the tires" and find out whether a provider is really worth committing to. Some are… some deadpool.
IMHO, the mark of a provider that I'm going to wind up using for production and sticking with for a long time is transparency and consistency about quality. OVH is a great example: they have some very cheap services, but they are very clear about the limitations and are thus able to deliver exactly what they promise. I'm paying $5.99 for a Kimsufi dedi with 1 GB of storage that I use for backups. Is that "low quality"? I'd say, no. It very consistently delivers the 100 mbps of bandwidth, the ATOM CPU, and the modest SATA hard drive that it promises. If it promised me more than that, and if I relied on those promises for, say, a web server, I'd be disappointed (and a fool!). But it's everything I need for a place to stick some off-site backups.
By contrast, SSDNodes is infamous here for overpromising tons of disk, bandwidth, etc., for a very low price (always, like, "90% off"). Sometimes, you actually get great performance from them; sometimes, it's garbage. That inconsistency means you can't rely on them and makes them "low quality" in my book (though I do still use them for a few things, always assuming that I will get only a fraction of the promised performance).
Thanks for your business
In a way, all the suggested options for real production (not just family and friends) stay in the $4-5/m range and those offers are 1/3 of the price so something is not "right".
@oplink is too low-key on LET. One of the most solid providers on my list. Zero down time for 1+ years and the Geekbench stats maintained the same level as the first day.
And @jar and his MXRoute will never fail you when you need a reliable email provider with optimized inbox delivery rate.
No offence. Is it really? Did they have the accident by hurricane some months ago? You mean a hardware not network?
We love low-key or underrated comments:) Reminds me of a good car from the factory who gives you more horsepower than advertised!
-Ryan
Our hardware is always great. We had some trouble with our uplinks during the last hurricane. Situation was out of our control and our two providers Lumen and Cogent said they have resolved/improved the power issues they had from the storm.
-Ryan
Another example is OVH. I had an old KS-1 for just under 10 years at £5 per month. That was before I was savvy enough to monitor network uptime, but the server itself never even experienced a single power failure in that time. At one point I had an uptime well over 1000 days because I forgot about upgrading the distro for a while.
My current Hetzner dedi has an uptime over 700 days (not too worried about security as the host has no public services available, SSH only over wireguard and the individual VMs which are port forwarded to the outside are regularly updated). According to hetrixtools, the network uptime is 99.9944%.
My current OVH dedi similarly has over a year of uptime, it only went down because the PSU failed, and that was detected by OVH's internal systems when I tried to use the control panel to reboot and it never came up, and they diagnosed the issue and replaced the PSU and sent me an email saying what they'd done within an hour or me trying to reboot it. Including that downtime, according to hetrixtools, I've had 99.9781% network uptime.
I have a few VPS with other providers, especially GreenCloud and Hosthatch. Never used them for a lot, but e.g. my GC vps in HK over the last 2 years has 99.9995% uptime, in SG 99.9408%, etc. My Racknerd in LA is at 99.3733% despite having an 8 hour network outage quite a few months ago.
All in all, I'd say that most of the established providers are very reliable. But I'd also suggest having redundancy across multiple providers whatever price point you use, because random failures can occur with any provider.
When I said "cutting corners" I did in no way imply that the service was bad, I am a very happy Greencloud customer myself. Just to make this perfectly clear, I am in no way trying to shame Greencloud, they deliver tremendous value for money and has never let me down.
What I meant was that there is a reason some vps's cost hundreds of dollars a month, while some costs $7 per year.
If you don't need 24/7 support by highly trained engineers, triple redundant tier 1 uplinks or mirrored datacenters in different geographical locations, then maybe the hundred dollars per month vps is not for you. If you're fine with a one-man show with a single pop that are dependent on other providers, then you will probably be totally happy with the $7.
Maybe "cutting corners" was a bad way to put it, what I meant was that there are differences in what providers sell. I do in fact pay hundreds of dollars a month for vps's for major projects, If a $7 per year could deliver exactly the same thing I would truly be an idiot. Not saying I'm not, but you get the point...
Do some providers make redundancy easier? Let's say I'm just starting out with shared hosting at a provider that is popular on LowEndTalk. Call it "BallsHost". Some months later, I finally get around to thinking about redundancy. The more I dig into the subject, I discover BallsHost lacks a crucial feature that would allow efficient redundancy. So I regret choosing BallsHost.
Is this scenario even remotely possible?
Like you've said, it was an accident by hurricane, while the uptime of my VMs kept continuously counting. Or, if you wish, I could change my words to "Zero down time (caused by the poor service management of the provider) for 1+ years".
God...
I am curious on what your thoughts would be if Aws/ Azure or gcp offered services at 7/ year.
Just hypothetical.
That comes down to how one wants to manage their infrastructure at the end and how important uptime is. You can have similar things possible with low cost players in market.
Over the years I have come to acknowledge and accept that high cost is not always premium. There are wanna be aspirants with at want to break the norm and disrupt and offer similar level of services.
In one of my work projects, the company had invested over 100M$ to Azure and still the response times / acknowledgement of issues by support took long time to acknowledge, let alone be fixed.
Just want to add, usually one pays for a Brand more than anything else.
I believe it's just perception at the end.
My 2 cents.
For instance, how I described it myself. All my HTTP endpoints on the various providers are running haproxy, which forward to multiple backends. You can have those backends on the same host, failover to a central server, operating on a cluster of servers, whatever.
Whether your backend can be distributed depends on what it is - if it's just static web assets, then you can easily replicate them to other machines however you want. If it's a web app, maybe you can have multiple servers talking to a shared (and maybe distributed) database.
In my case, because I'm writing the backend myself, I've architected it so that it only uses a local database, and new records are spooled back to a central server and then propagated back out to all nodes. The app front-end is also tolerant to lost records, and will retransmit its own records again after an hour if it hasn't received an acknowledgement, which only happens after the server nodes have received the data that was uploaded. It's more complicated that just setting up a big distributed database somewhere, but I designed it specifically so I can use small low-cost VPS instances.
None of that is dependant on any particular provider, although some providers provide additional features you might find really useful - e.g. BGP support so you can have a single IP address across multiple providers in many different regions (but this is expensive as you need a whole /24) or if you're happy to be tied to a single vendor, you could look at BuyVM which does the same themselves across 4 different regions and because they own the IP range, you're only paying for the 1 IP address you need.
Some of the proxy providers also do this kind of thing, so e.g. if you pay Cloudflare enough, you can have geo-located backends.
But this doesn't save you if the main server that's running haproxy goes down?
Multiple haproxy instances on different providers that each can talk to multiple backend servers. In my case, I use round-robin DNS so clients will try different ones after a couple of seconds if the first is down for some reason.
Good job!
Nice, I'm planning to try that out myself with 2 low end boxes. Currently trying to figure out how this will work with LetsEncrypt SSL certificates.
Been using @labze hostbrr in productiion for over 8 months now. Never heard any complaints from my client so far!